Model profile
Mistral Small 3.1
Evidence summary
Mistral Small 3.1 has an estimated overall rank of #171; its 90% source-sensitivity interval is #65–#251. Its behavior-only rank is #163; company governance moves the combined estimate to #171. Published evidence spans 4 evals and 2 of 7 behavior components. Its strongest relative result is BlueBench AttaQ-100 (attaq_harmlessness_reward_pct, #5 of 18); its weakest is Cisco AI Defense Rolling Single-Turn Leaderboard (single_turn_attack_success_rate, #96 of 105).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗mistral-small-3-1
- Hugging Face ↗mistralai/Mistral-Small-3.1-24B-Instruct-2503
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗repository created at proxy
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #128 / 311 | 0.7684 | ↓ lower | Source ↗official | |
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #5 / 18 | 87.62 | ↑ higher | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #96 / 105 | 74.33 | ↓ lower | Source ↗official | |
| PHAREhallucination_resistance_diagnostic | #44 / 70 | 0.7 | ↑ higher | Source ↗official | |
| PHAREharm_resistance_diagnostic | #47 / 70 | 0.9091 | ↑ higher | Source ↗official |
Values evaluations
Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.